caramoan tour package

caramoan tour package

caramoan tour package

Author Topic: QSC Amp Topology - Unravelling the Mystery  (Read 14226 times)

Offline tony

  • Technical People
  • Solar Power Satellite
  • *****
  • Posts: 7179
  • Pogi/Ganda Points: 668
  • Gender: Male
  • leach amp afficionado
Re: QSC Amp Topology - Unravelling the Mystery
« Reply #140 on: September 26, 2011, 01:06:05 PM »
^yes, since those caps are in parallel, then the esr's are also in parallel thereby reducing them effectively better than a single cap....
We are all LEARNERS here, strive to learn how circuits work and never let your subjective perceptions cloud your understanding of circuits....
http://www.diyaudio.com/forums/power-supplies/191730-tonys-latest-traffo-diy-build-12.html
http://ronelex.com/bias-current-output-transistor-amplifier/156/

Philippine Electronics Forum

Re: QSC Amp Topology - Unravelling the Mystery
« Reply #140 on: September 26, 2011, 01:06:05 PM »

Offline labgruppen

  • Lead Acid Battery
  • *******
  • Posts: 943
  • Pogi/Ganda Points: 103
Re: QSC Amp Topology - Unravelling the Mystery
« Reply #141 on: September 26, 2011, 01:10:10 PM »
Plus the ESR is inside the loop, feedback mechanism will compensate..

Philippine Electronics Forum

Re: QSC Amp Topology - Unravelling the Mystery
« Reply #141 on: September 26, 2011, 01:10:10 PM »

Offline TinTopHack

  • Lead Acid Battery
  • *******
  • Posts: 944
  • Pogi/Ganda Points: 155
  • Gender: Male
  • Nothing gold can stay. - Robert Frost
Re: QSC Amp Topology - Unravelling the Mystery
« Reply #142 on: September 26, 2011, 01:21:33 PM »
another secret to this amp would be that the filter caps must be of equal value, otherwise, the voltage developed across them will not be equal, thus contributing further to voltage offsets....i suppose you can install equal value resistors to force the issue...

ecaps are notoriously wide tolerance caps, matching would then be desirable....

this is perhaps the reason why another version used a center tapped power transformer....

My initial analysis of the global feedback loop, it appears to take care of insuring the imbalance is minimized if not eliminated. Since the output is taken from the junction of the 2 groups of ecaps and the feedback is taken from the same point, any imbalance that results in an output offset under no signal conditions, can be detected and corrected.

Sir in models with R14,19 paired with a trimmer, the trimmer is actually AC coupled to the junction of R14 and R19 via a capacitor. And in DC-coupled models of QSC, totally wala na ang R14 and R19.  
*DC-coupled models have a transformer center tap connected to the junction of C3,4 and C13,14 much like a conventional power supply. However, unlike the conventional power supply, it is  still not connected to ground.

You are right they are AC coupled. I did not see that at first. They are actually used for hum cancellation. I wil try to cover this in Part 3.

The use of transformers with center tap reduces the +/-VS sag and imbalance during operation at high power levels. BUT it complicates short circuit protection. Still +/-VS floats with respect to ground.
The world, as everybody knows, is analog; unless, of course, it's digitized.

Philippine Electronics Forum

Re: QSC Amp Topology - Unravelling the Mystery
« Reply #142 on: September 26, 2011, 01:21:33 PM »

Offline TinTopHack

  • Lead Acid Battery
  • *******
  • Posts: 944
  • Pogi/Ganda Points: 155
  • Gender: Male
  • Nothing gold can stay. - Robert Frost
Re: QSC Amp Topology - Unravelling the Mystery
« Reply #143 on: September 26, 2011, 06:05:56 PM »
Part 3: Input Stage + Output Stage

Below is a simplified diagram of the input stage combined with the output stage. The output stage is represented by an inverting amp with a gain of -A. This gain as shown in previous calc can be from 27 to almost a hundred in labgruppen's SIM. The input stage is just an op-amp. But the open-loop gain of an op-amp is typically around 100,000. Multiply that with the gain of the output stage of let's just say 30, we get an overall open-loop gain of 30 million!!!

In contrast, a conventional LIN amp, the VAS can have a voltage gain of around 500 to 800. Multiply this with the typical gain of the input diff stage of 300, you get an open loop gain of 150,000 to 240,000.



Since the output stage is inverting, the input op-amp is configured as an inverting amp so that whole amp is non-inverting. For this reason, the global feedback resistor R27 from the output connects to the non-inverting input of the op-amp to create negative feedback. As calculated by some of you before, the closed-loop gain is 1 + R27/R8 = 20.36

The global feedback also insures that the output voltage at no signal conditions is practically 0V. It is possible that the cap pairs C3/C4 and C13 /C14 are not equal in value and characteristics and therefore may not equally divide the rectified voltage from the transformer - resulting in +Vs not equal to -VS. To correct this problem, the voltage divider formed by R14 and R19 divides the rectified voltage equally and introduce this as a reference point into the op-amp. For this purpose, R14 and R19 must have 1% tolerances. any imbalance between +Vs and -VS will be automatically detected and corrected.
 
C6 and R29 is a form of freq compensation for the output stage to keep it stable.

Clipping Detection
The opamp and the output stage act as one big op-amp configured as a non-inverting power amp. During operation then there is no clipping the voltages at non-inverting input of the op-amp will track the voltage at the inverting input. This means due to the feedback mechanism in place the voltage at the non-inverting input and the voltage at the inverting input will always be equal. Remember the op-amp will always do whatever is necessary at its output to keep the voltages at its inputs equal

When clipping happens, the feedback loops detects that the output no longer tracks the input. For example, a positive going signal enters the input of the op-amp. The op-amp inverts this so its output will be negative going. The output stage inverts this so Vout at the speaker is positive going. Since the voltage gain of the whole amp is 20.36, Vout will try to be 20.36 times the input voltage. Clipping occurs if this is not possible. In this case, the positive going Vout can no longer go higher as needed to keep the voltage at the op-amp non-inverting input equal to the voltage at the inverting input. As a result the voltage at the inverting input will be higher than the voltage at the non-verting input. This condition, will cause the op-amp output to go low expecting that the voltage at the non-inverting input will rise. But this does not happen so its output will go lower until it is big enough to turn-on the clipping LED indicator. This was shown by labgruppen in his SIM waveforms. In fact if the input is overdriven, the output of the op-amp will even reach  +ve and -ve staturation - almost +/-15V.
 
The world, as everybody knows, is analog; unless, of course, it's digitized.

Philippine Electronics Forum

Re: QSC Amp Topology - Unravelling the Mystery
« Reply #143 on: September 26, 2011, 06:05:56 PM »

Offline TinTopHack

  • Lead Acid Battery
  • *******
  • Posts: 944
  • Pogi/Ganda Points: 155
  • Gender: Male
  • Nothing gold can stay. - Robert Frost
Re: QSC Amp Topology - Unravelling the Mystery
« Reply #144 on: September 26, 2011, 11:52:50 PM »
Correction: 30 million open loop gain above should be 3 million.
The world, as everybody knows, is analog; unless, of course, it's digitized.

Philippine Electronics Forum

Re: QSC Amp Topology - Unravelling the Mystery
« Reply #144 on: September 26, 2011, 11:52:50 PM »

Offline labgruppen

  • Lead Acid Battery
  • *******
  • Posts: 943
  • Pogi/Ganda Points: 103
Re: QSC Amp Topology - Unravelling the Mystery
« Reply #145 on: September 27, 2011, 12:14:16 AM »
QSC schematics..

AC-coupled models

RMX850 Click to View full size



USA400 Click to View full size



DC-coupled models

RMX1850HD Click to View full size



USA1310 Click to View full size


Notes:
1. All schematics are copyrighted by QSC Audio.
2. RMX series are current production models. USA series are now discontinued (only the series, not the topology.  ;) )
3. USA1310 has a relay for DC-fault Protection, present day models use crowbar circuit.
4. Present day models have now High Pass Filter options formed by the additional RC components at U201:2.. A clip limiter is also provided by the OTA U10:2.
5. It can also be noticed from USA series schematics that the muting/thermal protection circuit just shorts-out the op-amp supply rails. No op-amp rails, no drive signal to the output stage, no output signal. Neat!

BTW, I think another function of R14,19 is to close the DC-loop.. Without those, there is no feedback path for DC thus just minor DC input offset will make the rails become unbalanced.  And I agree that the center tap for the higher powered models will also force the voltages across the capacitors to equalize.  Lalo na ang real world music signals minsan hindi pareho ang magnitude ng + at - swings lalo na sa transients like kick drums and these transients demand lots of power.  

EDIT:
fixed preview image for USA1310 schematics.

Offline TinTopHack

  • Lead Acid Battery
  • *******
  • Posts: 944
  • Pogi/Ganda Points: 155
  • Gender: Male
  • Nothing gold can stay. - Robert Frost
Re: QSC Amp Topology - Unravelling the Mystery
« Reply #146 on: September 27, 2011, 04:46:29 AM »
Thanks to labgruppen for posting the schematics and his comments. +1
Question: When you say DC coupled models are you referring to those models with center-tapped transformers?

Part 3: Input Stage + Output Stage - Continued...

Below is the big op-amp representation of the input and output stage combined. Here it is clear how the global feedback loop sets the gain to approx. 20.  It also shows how the big-amp is configured as a non-inverting amp while the inner op-amp operates as an inverting amp. The additional inversion done by the output stage makes the whole thing a non-inverting amp.



There are previous postings about high open loop gain and using large amounts of feedback that said these are undesirable.

Let us see what do the experts say:

The late Randy Slone, in his book, "The Audiophiles Project Sourcebook" wrote: "High performance amplifiers require very high open-loop gains so that large quantities of negative feedback can be applied for optimum linearity response. In the general sense, the higher the open-loop gain the better the THD performance will be."

Horowitz, in his book, "The Art of Electronics"  wrote: "... effects of feedback. The most significant are predictability of gain and reduction of distortion...".

Douglas Self, in his book Audio Power amplifier Design Handbook 5th Ed., wrote the following:
"The main use of NFB in power amplifi ers is the reduction of harmonic distortion, the reduction of output impedance, and the enhancement of supply-rail rejection. There are also analogous improvements in frequency response and gain stability, and reductions in DC drift."

"Negative feedback is, however, capable of doing much more than stabilizing gain. Anything untoward happening in the amplifier, be it distortion or DC drift, or any of the other ills that electronics is prone to, is also reduced by the negative-feedback factor (NFB factor for short)."

"...the higher the open-loop gain A compared with the gain demanded by ß (negative feedback network) , the lower the distortion."


May I also refer you to Chapter 2 of the same book in the section entitled, "Some Common Misconceptions about Negative Feedback" for more of his viewpoints.

Audio opamp nothwithstanding, perhaps one other reasons why some audio designers still shy away from opamps is the fact that opamps has a very high open loop gain. This necessitates the use of heavy negative feedback to get it to do some useful (linear) functions.

Which is not bad at all at first glance. Negative feedback does a lot of wonderful things to an amplifier circuit, but it too has a bad side. Example of bad effects could be added instability. The good news is, instability can be easily tamed. Another bad effect is poorer transient response. In a negative feedback circuit, the fact that it takes a finite amount of time before the sampled output signal propagates back to the input means there will be a brief amount of time where the opamp “sees” no feedback at all. This phenomenon reveals itself as spikes in the output edges when the amplifier is fed with an input rectangular waveform. Minimizing this is the greater challenge.

That's why some audio manufacturers uses as little negative feedback as possible, and advertises this as a desirable feature.  ;D

While negative feedback does have its problems, there are solutions for that. I guess as long as the design achieves its purpose (design criteria) and the use of high open-loop gain and negative feedback helps achieve that purpose then that is fine with me.

The fact is active electronic devices are non-linear by nature. The use of high open-loop gains and negative feedback is currently the most effective way (in terms of cost and performance) of getting useful linearity out of them. Until somebody invents something better, we have to live with them somehow.
The world, as everybody knows, is analog; unless, of course, it's digitized.

Offline labgruppen

  • Lead Acid Battery
  • *******
  • Posts: 943
  • Pogi/Ganda Points: 103
Re: QSC Amp Topology - Unravelling the Mystery
« Reply #147 on: September 27, 2011, 06:01:52 AM »
Yes sir, the ones with center-tapped transformers are not AC-coupled, as such these models have DC-fault protection to prevent damage to speakers should a fault occur. With the center tap connected and if for example the positive rail output transistor is shorted, the upper capacitors will not be free to discharge to zero volt (it will be continuously charged up like in a conventional power supply) and the lower capacitors likewise will not be free to charge up rail to rail.




without center-tap, positive rail shorted. Shorted output transistor will just result to a one big thump and the output voltage will settle to almost zero if not zero. C13,14 charges up rail to rail (70V) and C3,4 disharges to zero. Speaker will most probably survive hence no DC-fault protection is placed at the output.


with center-tap, positive rail shorted. Shorted output transistor in one of the rails will result to a busted speaker if not equipped with DC-fault protection. Voltage across C3,4 and across C13,14 remains reasonably balanced. 32+Vdc appears across the load indefinitely if no protection scheme is used.






Offline Johannah Aleksandria

  • Gas Turbine
  • **
  • Posts: 2993
  • Pogi/Ganda Points: 163
  • Make it simple
Re: QSC Amp Topology - Unravelling the Mystery
« Reply #148 on: September 27, 2011, 06:22:48 AM »
Can anybody show the power computation.
What would be the rated power of the amp at +/- 50v?
Generations come and generations go, but the world stays just the same. What has happened before will happen again.

Offline labgruppen

  • Lead Acid Battery
  • *******
  • Posts: 943
  • Pogi/Ganda Points: 103
Re: QSC Amp Topology - Unravelling the Mystery
« Reply #149 on: September 27, 2011, 11:56:03 AM »
^the power computation should be the same as that of conventional topology amps. 100W is possible at 8R with +/-50Vdc  rails(with allowance for up to 10V drop on supply rails and power stage drops).

Anatomy of an Amplifier  <- An article written by Pat Quilter for S&VC magazine. Just an overview for amps.

Offline Johannah Aleksandria

  • Gas Turbine
  • **
  • Posts: 2993
  • Pogi/Ganda Points: 163
  • Make it simple
Re: QSC Amp Topology - Unravelling the Mystery
« Reply #150 on: September 27, 2011, 12:56:35 PM »
Sir Lab what is your take about the operation of the amp?
Is it the same with sir TTH?
Do you have any other explanation how the circuit works? Just the output stage.
Generations come and generations go, but the world stays just the same. What has happened before will happen again.

Offline Johannah Aleksandria

  • Gas Turbine
  • **
  • Posts: 2993
  • Pogi/Ganda Points: 163
  • Make it simple
Re: QSC Amp Topology - Unravelling the Mystery
« Reply #151 on: September 27, 2011, 02:55:47 PM »
There have been a few power amplifier designs incorporating voltage gain in the output stage. The primary motivation for this design approach has been the unavailability of high-quality high-voltage complementary output devices. Now that these are readily available, the technique is pointless and somewhat impractical.

A quote from Randy Slone from his book "High Power Audio Amplifier Construction Manual".
Generations come and generations go, but the world stays just the same. What has happened before will happen again.

Offline TinTopHack

  • Lead Acid Battery
  • *******
  • Posts: 944
  • Pogi/Ganda Points: 155
  • Gender: Male
  • Nothing gold can stay. - Robert Frost
Re: QSC Amp Topology - Unravelling the Mystery
« Reply #152 on: September 27, 2011, 04:48:30 PM »
Clarification:

In my simplified diagrams above, the output stage refers to the combination of the 2 transistor stages:
1. The driver stage compose of Q2 and Q3 which functions like a VAS in the conventional LIN topology
2. The final output stage compose of Q1 and Q4 which are emitter followers and therefore has a voltage gain of  approx.1.

We are used to the LIN topology where the final output stage and driver transistor stage work together as one high current gain (hfe) stage - with a voltage gain of approx. 1 - because they act as emitter followers. We might easily treat the driver stage of QSC topology as the same as the driver stage of LIN topology. BUT they are not the same. For the same reason the combination of Q2 and Q1 in the QSC topology MUST NOT be  treated as a Sziklai pair. As I pointed out they are separate stages. Q2 is a common emitter stage with voltage gain and Q1 is an emitter follower stage.

The statement of Randy Slone in his book "High Power Audio Amplifier Construction Manual" refers to the LIN topology on which all of his designs were based.
The world, as everybody knows, is analog; unless, of course, it's digitized.

Offline TinTopHack

  • Lead Acid Battery
  • *******
  • Posts: 944
  • Pogi/Ganda Points: 155
  • Gender: Male
  • Nothing gold can stay. - Robert Frost
Re: QSC Amp Topology - Unravelling the Mystery
« Reply #153 on: September 27, 2011, 08:43:28 PM »
AC Coupling or DC Coupling?

Labgruppen's categorizing the QSC amps models into AC coupled and DC coupled is quite interesting. I believe this is a good topic for discussion. Let us take a look at the 2 categories:


It is interesting to note that the current paths when the upper half of both circuits conduct are the same. The only difference is the circuit without center tap connection will have a bigger tendency to have the voltage across C3/C4 sag because it has to replenish the charge being dumped through the speaker load via C13/C4. The circuit with center tap connection has a direct path for replenishing the charge being dumped through the speaker.

It is also interesting to note that as far as the rest of the circuit is concerned, C3/C4 is just a very low impedance voltage source or a battery - irregardless of where the current used to charge it comes from.

As long as the circuit operates normally, the operations are practically identical. Only when an output transistor gets shorted that a difference comes out. Under this condition, the circuit with center tap connection will continue to supply current to the short circuit section in the form of pulsating DC because C3/C4 can no longer charge up to the nominal +VS value.

Now, let us compare the current paths of these 2 circuits to the current path in a conventional output stage of a LIN topology amp as shown below. Here, the load is generally considered as DC coupled. Do you see any difference in current path? Let's also compare them to the current path of what is generally accepted as an AC coupled amp.


From this comparison, how do we categorize the QSC amp models?
The world, as everybody knows, is analog; unless, of course, it's digitized.

Offline labgruppen

  • Lead Acid Battery
  • *******
  • Posts: 943
  • Pogi/Ganda Points: 103
Re: QSC Amp Topology - Unravelling the Mystery
« Reply #154 on: September 27, 2011, 08:54:40 PM »
AC Coupling or DC Coupling?

Labgruppen's categorizing the QSC amps models into AC coupled and DC coupled is quite interesting. I believe this is a good topic for discussion. Let us take a look at the 2 categories:


It is interesting to note that the current paths when the upper half of both circuits conduct are the same. The only difference is the circuit without center tap connection will have a bigger tendency to have the voltage across C3/C4 sag because it has to replenish the charge being dumped through the speaker load via C13/C4. The circuit with center tap connection has a direct path for replenishing the charge being dumped through the speaker.

It is also interesting to note that as far as the rest of the circuit is concerned, C3/C4 is just a very low impedance voltage source or a battery - irregardless of where the current used to charge it comes from.

As long as the circuit operates normally, the operations are practically identical. Only when an output transistor gets shorted that a difference comes out. Under this condition, the circuit with center tap connection will continue to supply current to the short circuit section in the form of pulsating DC because C3/C4 can no longer charge up to the nominal +VS value.

Now, let us compare the current paths of these 2 circuits to the current path in a conventional output stage of a LIN topology amp as shown below. Here, the load is generally considered as DC coupled. Do you see any difference in current path? Let's also compare them to the current path of what is generally accepted as an AC coupled amp.


From this comparison, how do we categorize the QSC amp models?
Well sir, regarding DC/AC coupling, IIRC, I think I read it from a QSC service manual.  :)   I just can't find it this time, thus I just posted the simulation to prove it is DC coupled.  To further simplify and avoid the complications of the bridge rectifier and transformer, we can replace them with actual DC source, batteries.

Offline labgruppen

  • Lead Acid Battery
  • *******
  • Posts: 943
  • Pogi/Ganda Points: 103
Re: QSC Amp Topology - Unravelling the Mystery
« Reply #155 on: September 27, 2011, 11:32:45 PM »
Sir Lab what is your take about the operation of the amp?
Is it the same with sir TTH?
Do you have any other explanation how the circuit works? Just the output stage.

Well I already posted 3days ago my understanding on how the output stage works.. just a very simple explanation.
... The input signal is fed to its inverting input (U1B). Thus a positive going signal will make U1B's output go negative. This negative going signal makes Q3 conduct which in turn makes Q4 conduct. (When Q3 conducts, its collector pulls UP the base of Q4 and makes Q4 conduct) When Q4 conducts this will pull the negative rail up towards ground. Since the power supply is floating(not grounded), both the positive and negative rails will go up. This will make the output (center tap of C3,4 and C13/14) which was initially at ground level go higher that ground, thus a positive going output. The reverse is true for negative going signals.
..

Too bad, the service manual of QSC doesn't explain the operation down to the teeniest bit.




Uploaded with ImageShack.us


Offline labgruppen

  • Lead Acid Battery
  • *******
  • Posts: 943
  • Pogi/Ganda Points: 103

Offline TinTopHack

  • Lead Acid Battery
  • *******
  • Posts: 944
  • Pogi/Ganda Points: 155
  • Gender: Male
  • Nothing gold can stay. - Robert Frost
Re: QSC Amp Topology - Unravelling the Mystery
« Reply #157 on: September 28, 2011, 12:36:21 AM »
Thank you Labgruppen. This explanation perfectly match the illustration I made to explain how the output stage works in Part 1. I just did not have the right words to explain it. In fact your explanation also match my illustration.

If we combine the 2 drawings I made in Part 1, we get this:


This matches the QSC explanation:
- that common emitter circuits drive the supply rails - emitter followers
- that devices drives the rails with audio signals which rides atop the DC
- the collector of each driver transistor directly drives the base of its output transistors - Thus it is correct NOT to treat them as Sziklai pairs.

The biasing which I am supposed to discussed next is also explained. But the details on current limiting is not explained in detail. That I will try to expound on.



The world, as everybody knows, is analog; unless, of course, it's digitized.

Offline labgruppen

  • Lead Acid Battery
  • *******
  • Posts: 943
  • Pogi/Ganda Points: 103
Re: QSC Amp Topology - Unravelling the Mystery
« Reply #158 on: September 28, 2011, 02:36:13 AM »
Just a silly question sir TTH, if I use a sziklai pair as a switch, will it also be correct not to treat it as a sziklai pair? (Just to clarify, I understand your explanation and I am not contesting it.  In fact when I read it, I thought "oo nga ano..")  
Now, I just had a musing that I can also apply the same explanation in sziklai pairs used as a switch. Example below, can I say that Q2 is like a VAS with a very high voltage gain (no Re + very high effective collector resistance), Q1 an EF for Q2 since Q1's emitter follows the voltage at the collector of Q2) thus not treating them as a sziklai pair?  Now question again, when shall we treat a sziklai pair as sziklai pair? Because when we see the diagram of a sziklai pair, we can say that the right-side transistor will always be an emitter follower for the collector of the left-side transistor. Though this explanation can also explain why sziklai pairs have very high current gains. Since Q1 is acting like an EF for Q2, this means that Q2 sees only a very small fraction of the load current (ILOAD/BQ1). Thus Q2 requires only a very minimal base current to saturate ((ILOAD/BQ1) / BQ2). Thus the current gain of sziklai pair is BQ1*BQ2..


EDIT:
BQ1  = Beta Q1
BQ2  = Beta Q2







Offline TinTopHack

  • Lead Acid Battery
  • *******
  • Posts: 944
  • Pogi/Ganda Points: 155
  • Gender: Male
  • Nothing gold can stay. - Robert Frost
Re: QSC Amp Topology - Unravelling the Mystery
« Reply #159 on: September 28, 2011, 04:40:12 AM »
Just a silly question sir TTH, if I use a sziklai pair as a switch, will it also be correct not to treat it as a sziklai pair? (Just to clarify, I understand your explanation and I am not contesting it.  In fact when I read it, I thought "oo nga ano..") 
Now, I just had a musing that I can also apply the same explanation in sziklai pairs used as a switch. Example below, can I say that Q2 is like a VAS with a very high voltage gain (no Re + very high effective collector resistance), Q1 an EF for Q2 since Q1's emitter follows the voltage at the collector of Q2) thus not treating them as a sziklai pair?  Now question again, when shall we treat a sziklai pair as sziklai pair? Because when we see the diagram of a sziklai pair, we can say that the right-side transistor will always be an emitter follower for the collector of the left-side transistor. Though this explanation can also explain why sziklai pairs have very high current gains. Since Q1 is acting like an EF for Q2, this means that Q2 sees only a very small fraction of the load current (ILOAD/BQ1). Thus Q2 requires only a very minimal base current to saturate ((ILOAD/BQ1) / BQ2). Thus the current gain of sziklai pair is BQ1*BQ2..


EDIT:
BQ1  = Beta Q1
BQ2  = Beta Q2


A Sziklai pair is a replacement for a darlington pair - the purpose of which is to achieve a very high current gain. So if a darlington pair can be used as a switch then a Sziklai pair can also be used as a switch. So if you use a Sziklai pair as a switch then there is no reason why you should not treat it as a Sziklai pair. Treating it as a Sziklai pair in a switching application will simplify your job. You can just think of it as one transistor with a gain of Beta1 x Beta2.

When do you treat a Sziklai pair as a Sziklai pair? I think the question should be: When do you treat a circuit that looks like a Sziklai pair as a Sziklai pair? First step is to find out what the circuit does. If the compound circuit (that looks like a Sziklai pair) can be treated as one transistor to do its intended function then you can treat it as a Sziklai pair.

In the case of the QSC output stage that we discussed here, there are two fundamental signs that it should not be treated as a Sziklai pair. The first transistor has an emitter (degenerative) resistor and the second transistor also has an emitter resistor. The Sziklai pair is intended to work as one single transistor with very high gain. As such you would not put 2 degenerative resistors in the circuit. Instead, if you really need it, you just need one and it should be connected to the junction of the emitter of the first transistor and the collector of the second. 
   

 
The world, as everybody knows, is analog; unless, of course, it's digitized.

Philippine Electronics Forum

Re: QSC Amp Topology - Unravelling the Mystery
« Reply #159 on: September 28, 2011, 04:40:12 AM »

 

Privacy Policy

Contact Us: elabph@yahoo.com